PROJECT SUMMARY/ ABSTRACT Classically, memory T cells arise after a productive immune response to a foreign pathogen in the periphery, and are poised to respond more rapidly to repeated pathogen challenge. However, in mice that have never encountered foreign pathogens, there is a major subset of endogenous CD8+ T cells that express phenotypic markers of immunological memory, making up 10-20% of total peripheral CD8+ T cells. Despite the prevalence of this ?memory phenotype? CD8+ T cell (CD8-MP) population, the nature of the antigens recognized by these cells and the anatomical sites in which they develop remain incompletely defined. Previous work has been limited to the study of bulk polyclonal CD8-MP cell populations, which may exhibit phenotypic and functional heterogeneity. This proposal seeks to elucidate fundamental aspects of the biology of CD8-MP cells at the clonal level. Using a unique clonal approach, this work seeks to elucidate the origin, antigen specificity, and function of CD8-MP cells at steady state and in the context of oncogene-driven prostate tumors. The central hypothesis of this proposal is that TCR recognition of endogenous self-ligands in the periphery drives the differentiation of CD8-MP clones, as well as the selective recruitment of these clonotypes into murine prostate tumors. The studies in Aim 1 will determine the extent to which CD8-MP differentiation is driven by TCR recognition of endogenous self-ligands and to elucidate whether CD8-MP differentiation occurs in the thymus or in the periphery. These studies are expected to demonstrate that a large fraction of CD8-MP cells express TCRs that recognize self-ligands presented by classical MHC class-I molecules in the periphery, which drives their differentiation into the CD8-MP subset. The studies in Aim 2 will define the contribution of CD8-MP cells to the pool of tumor-infiltrating lymphocytes and determine the functional impact of CD8-MP cells on TRAMP prostate tumors. It is anticipated that CD8-MP clonotypes make up a substantial fraction of prostate tumor- infiltrating CD8+ T cells, suggesting that CD8-MP T cells reactive to non-mutated self-ligands may play a unique functional role in the tumor environment. Such evidence would help explain a long-standing conundrum in tumor immunology, in which T cells reactive to tumor-expressed neo-antigens, differentiation antigens, or cancer testis antigens can be detected in primary human tumors but are typically rare, indicating that the majority of tumor-infiltrating T cells have undefined specificities. Ultimately, understanding the biology of this prominent subset of CD8+ cells at homeostasis and in the tumor microenvironment will provide unique insight for future preclinical studies aimed at manipulating autologous CD8-MP cells for therapeutic benefit.